Prediction league scoring systems, and which one stays competitive
Dardan Hasku3 min read
Every prediction league lives or dies on its scoring. Get it right and the table stays worth watching in March. Get it wrong and one person is uncatchable by Halloween, or worse, nobody can tell who is actually good.
The four systems people use
| System | Typical points | What goes wrong |
|---|---|---|
| Result only | 3 for the winner | Half the league ties every week |
| Result plus exact | 3 outcome, 5 exact | Exact scores are rare, so it is mostly a result league |
| Layered credit | Outcome, margin, goals, exact | Needs slightly more explaining |
| Layered plus multipliers | As above, times match difficulty | The most work to build, the most fun to play |
Almost every long-running league drifts down that table over time, because each row solves the problem the row above it creates.
Why result-only leagues go flat
In a normal weekend, most people pick most of the favourites. A result-only league therefore produces a cluster: five players on 18, three on 15, and no way to tell them apart. Ties are not a scoring detail, they are the whole experience of that format.
Adding a bonus for the exact score helps less than it looks, because exact scores land about one match in ten. Most weeks, nobody in a small league hits one, and you are back to a result league with extra steps.
Layered credit, and sensible numbers
Layering means paying separately for each thing you got right, so being nearly right is worth something. A set of values that works:
- Correct outcome: 20. The floor. Everyone who read the game right gets this.
- Correct goal difference: 10. Rewards understanding how comfortable the win would be.
- Each team's goals exactly right: 10 apiece. Two separate awards, so half-right calls score.
- Exact score: 10 on top. The bonus for putting the whole thing together.
The reason those specific numbers work is the spread they produce. A perfect call is worth 60, a good-but-wrong call is worth 20 to 40, and a miss is zero. Nobody is ever mathematically out of a round, and nobody wins one by accident.
Multipliers: the part that keeps it alive
Flat scoring still has a problem: calling the league leaders to beat the bottom club pays exactly as much as calling a genuine coin flip. So the safest possible coupon is the optimal coupon, and the league becomes a test of caution.
Two mechanics fix that, and they are the two worth stealing for any league:
- Difficulty multipliers. Price each match between about 1.0 and 2.5 based on how hard it is to call, and multiply the base points by it. Now bravery pays, and the safe coupon stops winning. How we set ours.
- A rationed double or nothing. One match a round where you back yourself: double if it lands, zero if it misses, with a monthly budget so you cannot use it every week. How the banker works.
Together they mean someone forty points behind has a real path back, without the league turning into a coin-flipping contest. That is the balance you are aiming for.
Two details worth writing down
- A round table as well as a season table. Weekly winners keep people who cannot win the season interested.
- Tiebreakers, agreed in advance. Most exact scores, then most correct outcomes, then the single biggest match score. Never leave it to be argued about after the fact.
All of the above is already running here, scored automatically: layered points, a difficulty stamp on every match, a rationed banker, and both a weekly and a season table.
See the full scoring